Narrative AI

“Hey, I’m going to subscribe to the universal library. Then, I’ll have all the future volumes of my periodical over and done with, already in the printer’s copy. I won’t need to worry about any of the articles. This is truly magnificent for the publisher: the elimination of the author from the entire business! The replacement of the writer by the combinatorial machine! A triumph of technology!” (Laβwitz, 1904)

It’s 1904. Editor Max Burkel and professor Wallhausen reflect on the opportunities that technology offers to publishing. Eased by a glass of Kulmbacher, a delicious Bavarian beer produced since 1895, their conversation explores the mathematical models needed to realize the project.

Kurd Laβwitz’s short story, The Universal Library, probes the idea of a library containing everything humankind knows and could know in the future. Considered by many the father of German sci-fi, Laβwitz launched a provocation, which sparked other writers’ creativity throughout the twentieth century. For instance, Jorge Luis Borges’s Library of Babel (1941) describes a universe as an infinite library holding all possible volumes with a specific format and character set.

Clockwise, from the top left corner: Italo Calvino, Umberto Eco, Jorge Luis Borges, Kurd Laβwitz.

The idea of a mathematical model able to (re)produce and organize all human knowledge has been really popular during the last century. And not only within speculative fiction but also as an organization and business problem.

Narrative Machines

NaNoGenMo (National Novel Generation Month) was a literary contest for computer-generated narratives. On the contest’s website, you can read all entries from 2013 to 2018. You just need to register an account on GitHub, but that’s easy and free.

Darius Kazemi launched the initiative on Twitter in 2013, and during the following five years it attracted numerous writers/developers.


One of the authors who participated in the contest, Andrew Plotkin, released the software he created, which wrote a 50K-word novel. Beyond the power of social media to foster such initiatives, these projects ask us to ponder what roles and opportunities human creativity has at the time of total digitization.

“Training” algorithms to produce literature has become popular. So much so that in 2017, former Obama administration’s ghostwriter Ross Goodwin connected a laptop, a camera, a GPS navigator, a microphone, and a mini-printer to write a novel in the style of Jack Kerouac’s On the Road. The laptop collects the data coming in from the other devices, mixes it with the hundreds of books Goodwin fed in its neural networks, and sends everything to the printer. Which churns out the novel. Produced and released by French publisher Jean Boîte Éditions, 1 The Road is not the first novel entirely written by an algorithm, but it’s the first travel story composed this way. However, can we imagine algorithms able to create works like Bruce Chatwin’s In Patagonia or Claudio Magris’s Danube? So, who are authors and what do they do?

Human-Machine Collaboration


With the Wob 2.0 revolution, participation and collaboration are keywords. So, in 2016, Myles O’Neill, Anthony Voutas, Isadora Lamego, and Annette Zouhanno released, a platform where people can create, share, and read artificial fiction. Leaving aside (for now) the pleonasm in its name, the project offers users the opportunity to both read other machines’ creations and to learn how to train their own algorithms to produce literary works.

For those who wonder in more specialized literary fields, there’s, a horror-dedicated platform. Inspired by Frankenstein (1818), this algorithm grabbed up all the stories collected on r/nosleep. Moreover, the website promotes human-machine collaboration. On its Twitter page, it publishes horror tales mixing readers’ nightmares and data elaborated by the algorithm.


There seems to be an undetected irony behind’s creation. The idea of reducing to a minimum—and eventually removing it—humankind’s role in the creative process already represents a sci-fi horror story for more conservative readers.

However, given these initiatives’ popularity, someone thought of launching a literary magazine almost entirely produced by artificial intelligence. The editors of CuratedAI—as the publication’s titled—describe it as “a literary magazine with a twist” and make sure that “Editing, for now, is still the domain of us humans, but we aim to keep our touch as light as possible.”




Despite these editors’ reassurances, these projects beg the same question raised by The Guardian, “Is the writing on the wall for human novelists?” Journalist Steven Poole asks if an algorithm-generated novel can have the same aesthetic value as Jane Austen’s or George Orwell’s works. Rather, I believe we should ask ourselves: what is the role authors, editors, and readers will have in a machine-dominated future?

Technology and Knowledge

Let’s think about our everyday language. It’s a means to make sense of the world around us and, sometimes, to modify it (for instance, let’s consider how some ritual phrases like “I baptize you” or “We have fully admitted XYZ to the degree of …” modify people’s identities). When we take into account the language’s sense-making and reality-molding capacity, we can conceive of it as a unique type of technology: an intellectual technology.

Every day, people use different languages to go about their businesses. For instance, they use maths and logic languages, which are the same codes used by computers to perform their tasks. Since machines are ubiquitous nowadays, we, as readers, editors, and writers, deeply experience with them a “space of computation” where we share “our most personal memories” and accept “algorithmic guidance on matters ranging from love to real estate.” (Finn, 2017)

The almost boundless presence of digital technologies pushes us to further consider the roles of the many players involved in cultural production and distribution. While pondering the “retraining of people in a completely automatized or strongly industrialized society,” Umberto Eco highlights that each time a new technology provides new opportunities to organize the world, it is necessary people learn how to use it.  Therefore, when we try to understand the roles of authors, editors, and readers in the digital world, Eco suggests that

“A discourse about what it is to be human does not lose its value or meaning: however, humans won’t be considered as syllogizing animals anymore, but as animals able to both build syllogizing machines and wrestle with new (original) problems about their use. […] And this process requires of intellectuals they take an adaptation class, eat the humble pie, and go back to school. ” (Eco, 286, my translation)

At the end of Laβwitz’s short story, with a refill of Kulmbacher, professor Wallhausen realizes its project’s limits and admits, “We need not search through the universal library for the volume we require, but rather create it ourselves in constant, serious, honest work.” Therefore, following Eco’s thoughts, our “constant, serious, honest,” creative endeavors will have to focus on different levels of abstraction and deal with tools, which may seem different, but are just languages after all.

Message me with your thoughts, or get in touch with me if you have an editorial/writing project you want to publish! See you next week.

This post’s contents, including its images, are published under a CC BY-NC license.


1 The Road. Jean Boîte Éditions.2017.

Casalegno, Giovanni. Storie di libri. Torino: Einaudi, 2011.

Eco, Umberto. Apocalittici e integrati. Milano: Bompiani, 1964.

Lasswitz, Kurd. The Universal Library. (1904)

Merchant, Brian. “When an AI Goes Full Jack Kerouac.” The Atlantic, October 1, 2018.

Poole, Steven. “The rise of robot authors: is the writing on the wall for human novelists?” The Guardian, March 25, 2019.

Racter. The Policeman’s Beard Is Half Constructed. 1984.

Finn, Ed. What Algorithms Want: Imagination in the Age of Computing. Boston (MA): The MIT Press, 2017. (Kindle Edition.)

One thought on “Narrative AI

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s